OpenAI and Broadcom announced the unveiling of Jalapeño, OpenAI’s first Intelligence Processor. Jalapeño is an AI accelerator designed for large language model inference and is the first chip in a multi-generation compute platform being developed by OpenAI and Broadcom.
The companies said the platform is intended to make advanced AI faster, more reliable, and more accessible.
Jalapeño was delivered to OpenAI CEO Sam Altman and President Greg Brockman by Broadcom President and CEO Hock Tan and President Charlie Kawwas.
OpenAI designed the chip from scratch around its understanding of LLM inference, informed by its roadmap across models, kernels, serving systems, and product needs.
Broadcom and Celestica helped industrialize the platform through chip implementation, board and rack system integration, high-performance networking, and scalable production systems.
The companies said Jalapeño is designed to work with current and future LLMs across the industry.
Engineering samples of the chip are running machine learning workloads in the lab at production target frequency and power.
OpenAI said early testing shows that Jalapeño is expected to deliver performance per watt substantially better than current state-of-the-art systems.
The architecture is designed to reduce data movement while balancing compute, memory, and networking resources to improve realized utilization closer to theoretical peak performance.
Broadcom’s silicon implementation and networking technologies, including Tomahawk networking silicon, are expected to support large-scale production and deployment.
OpenAI said Jalapeño is part of its broader full-stack infrastructure strategy, extending from products and models to chip architecture, kernels, memory systems, networking, scheduling, deployment systems, and user experience.
The chip was co-developed from initial design to manufacturing tape-out in nine months. OpenAI said the development process was accelerated by software-hardware collaboration and the use of OpenAI models in parts of the design and optimization process.
Jalapeño is expected to begin initial deployment by the end of 2026 and expand in the years ahead with data center partners.
The companies said the platform is intended to support gigawatt-scale data center deployments over multiple generations.
OpenAI said the goal is to improve the cost, speed, and reliability of inference, which can translate into faster ChatGPT responses, more capable Codex workflows, lower-cost API products, and more dependable access during high-demand periods.
KEY QUOTES:
“The world is moving to a compute-powered economy. Jalapeño is part of our long-term full-stack infrastructure strategy to make compute more abundant, resulting in AI which is faster, more reliable, more affordable for people and businesses, and can be used to solve more important problems. By designing more of the stack ourselves, we can serve more intelligence with greater efficiency and keep pushing advanced AI toward broader access.”
Greg Brockman, President and Co-Founder of OpenAI
“Jalapeño was designed from the ground up for LLM inference using detailed insights from our close collaboration with OpenAI researchers. We optimized the architecture around the kernels, memory movement, networking, and serving patterns that matter most for frontier AI models. Based on early testing, Jalapeño will efficiently execute our most important workloads close to the hardware’s theoretical limits.”
Richard Ho, Hardware Program Lead at OpenAI
“Our collaboration with OpenAI represents a fundamental commitment to scaling the physical infrastructure required for the next decade of AI. This is just the beginning of a multi-generation roadmap. By co-developing our industry-leading silicon directly with OpenAI, we are enabling the deployment of gigawatt scale data centers with Microsoft and other partners beginning in 2026.”
Hock Tan, President and CEO of Broadcom
